[159496] in North American Network Operators' Group
Re: OOB core router connectivity wish list
daemon@ATHENA.MIT.EDU (Dobbins, Roland)
Sat Jan 12 11:39:26 2013
From: "Dobbins, Roland" <rdobbins@arbor.net>
To: NANOG list <nanog@nanog.org>
Date: Sat, 12 Jan 2013 16:39:08 +0000
In-Reply-To: <CAL9jLabcyTQns31FO1Pys0Mc3-E_=Ng46Saj5njL1JwYmUWRQg@mail.gmail.com>
Errors-To: nanog-bounces+nanog.discuss=bloom-picayune.mit.edu@nanog.org
On Jan 10, 2013, at 9:35 AM, Christopher Morrow wrote:
> I don't think roland was really saying that normal netflow from a device =
in production pushing a few hundred gbps of traffic would be
> appropriate to ship out the OOB network... or I hope that wasn't his poin=
t. I don't think oob networks need to be sized for that.
Actually, that is what I'm saying - and with sampled flow telemetry, this i=
sn't a huge bps issue. With flow telemetry, we're typically looking at 0.5=
% - 1.5% of the aggregate utilized bandwidth of the exporting device, multi=
plied by the sampling ratio (e.g., .01, .0001, whatever) - so, they isn't a=
huge amount of traffic, even for very high-speed network.
Modern networks should be designed with OOB/DCN which are sufficiently size=
d to handle flow telemetry. Exporting it in-band means that one is blind d=
uring network partition events, when visibility is key.
-----------------------------------------------------------------------
Roland Dobbins <rdobbins@arbor.net> // <http://www.arbornetworks.com>
Luck is the residue of opportunity and design.
-- John Milton
